Phenotypic characterization of the CRISPA (ARP gene) mutant of pea (Pisum sativum; Fabaceae): a reevaluation.

American Journal of Botany 101(3): 408–427. 2014.

PHENOTYPIC CHARACTERIZATION OF THE CRISPA (ARP GENE) MUTANT OF PEA (PISUM SATIVUM; FABACEAE): A REEVALUATION1 DARLEEN A. DEMASON2 AND VENKATESWARI CHETTY Botany and Plant Sciences, University of California, Riverside, California, 92521 USA • Premise of the study: Leaf form and development are controlled genetically. The ARP genes encode MYB transcription factors that interact with Class 1 KNOX genes in a regulatory module that controls meristem-leaf determinations and is highly conserved in plants. ARP loss of function alleles and subsequent KNOX1 overexpression cause many unusual leaf phenotypes including loss or partial loss of the ability to produce a lamina and production of “knots” on leaf blades. CRISPA (CRI) is the ARP gene in pea, and a number of its mutant alleles are known. • Methods: We made morphological and anatomical evaluations of cri-1 mutant plants while controlling for genetic background and for heteroblastic effects, and we used aldehyde fixation and resin preparations for anatomical analysis. Further, we compared gene expression in WT and cri-1 shoot tips and HOP1/PsKN1 and CRI expression in other leaf mutants. • Key results: The cri-1 plants had more extensive abnormalities in the proximal than in the distal regions of the leaf, including ectopic stipules, narrow leaflets, and shortened petioles with excessive adaxial expansion. “Knots” were morphologically and anatomically variable but consisted of vascularized out-pocketing of the adaxial leaflet surface. HOP1/PsKN1 and UNI mRNA levels were higher in cri-1 shoot tips, and some auxin-regulated genes were lower. Low LE expression suggests that the GA level is high in cri-1 shoot tips. • Conclusions: The CRISPA gene of pea suppresses KNOX1 genes and UNI and functions to (1) maintain proximal-distal regions in their appropriate positions, (2) restrict excessive adaxial cell proliferation, and (3) promote laminar expansion. Key words: UNIFOLIATA.

adaxial meristem; ARP genes; compound leaves; CRISPA; Fabaceae; knots; leaf development; Pisum sativum;

Of the three vegetative organs of the plant body, only the leaf is typically dorsiventrally flattened, which must result from asymmetric gene expression and development. Leaves have three planes of asymmetry: proximodistal, adaxial–abaxial and mediolateral (Steeves and Sussex, 1989; Cronk, 2009). Leaves are also the most morphologically diverse vegetative organ on the plant body. Leaves can be simple, lobed, or compound, have toothed or smooth margins, possess or lack petioles and/or stipules, and have many variations in shapes and venation patterns. In addition, leaves have evolved morphological modifications to perform additional functions including tendrils for support, spines for protection, bladders for flotation, and traps for consuming insects. Now that a basic developmental framework for understanding morphogenesis of simple, bifacial leaves in Arabidopsis has been established, it is feasible to attempt to understand differences in these processes that explain morphogenesis of form variations in other

species and to evaluate these variations in development within a comparative context. Traditionally, we have described leaf morphogenesis as resulting from a series of transient meristems that function simultaneously and sequentially: the apical, adaxial, marginal, plate, and intercalary meristems (Cutter, 1971). The great variation in leaf form was attributed to the timing, relative activity, and duration of activity of these meristems (Cutter, 1971). An example of this is the adaxial meristem. This meristem was first recognized as a strip of cells below the adaxial protoderm, which divides periclinally (Foster, 1936). In simple, bifacial leaves this meristem functions solely to thicken the petiole–midrib axis, but in unifacial leaves, like circular leaves of onion, rachis leaves of the Apiaceae, ensiform leaves of Iris or Acorus, phyllodes of Acacia, peltate leaves, and tubular or epiascidiate leaves of insectivorous plants, it is the most important meristem controlling blade morphogenesis (Franck, 1976; Kaplan, 1970a, b, 1973a, b, 1975, 1980). Currently, we describe leaf morphogenesis as resulting from the activities of families of transcription factors. Among the most important and conserved of these are the mutually antagonistic Class I KNOX (KNOX1) and ARP (ASYMMETRIC LEAVES1, ROUGHSHEATH2, Phantastica) gene families, which function to define shoot apex vs. leaf. KNOX1 is expressed throughout the shoot apical meristem, where it negatively regulates ARP, reduces gibberellic acid (GA) accumulation, and is downregulated in the incipient leaf primordium (Hay et al., 2004; Piazza et al., 2005). This downregulation of KNOX1 allows ARP gene expression and an increase in GA concentrations within the leaf founder cells and in subsequent developing leaf primordia (Hay et al., 2004; Piazza et al., 2005). KNOX1 genes have been proposed

1 Manuscript received: 19 November 2013; revision accepted 27 January 2014. The authors thank Drs. Neil McHale, Patricia Polowick, and Patricia Springer for critical readings of the manuscript. This work was supported by the California Agricultural Experiment Station. The authors dedicate this paper to the memory of Donald R. Kaplan (17 January 1938–17 December 2007), whose mentorship of the first author (D.A.D.) made this work possible. Don was recognized for his rigorous analytical approach to plant form that allowed him to identify fundamental structural and developmental commonalities that transcended taxonomic boundaries. 2 Author for correspondence (e-mail: [email protected])

doi:10.3732/ajb.1300415

American Journal of Botany 101(3): 408–427, 2014; http://www.amjbot.org/ © 2014 Botanical Society of America

408

March 2014]

DEMASON AND CHETTY—REEVALUATION OF THE CRISPA PHENOTYPE OF PEA

to play the primary regulatory role in the development of compound leaves in many species of flowering plants (Bharathan and Sinha, 2001; Bharathan et al., 2002; Champagne and Sinha, 2004; Hay and Tsiantis, 2009). Overexpression of KNOX1 genes in tomato leaf primordia results in the production of very highly dissected compound leaves (Hareven et al., 1996; Chen et al., 1997; Efroni et al., 2010). Consistent with the fact that KNOX1 reduces GA levels, GA application reduces the level of branching in KNOX1 overexpressing tomato leaves (Hay et al., 2002, 2004). This demonstrates a negative relationship between GA levels and leaf complexity. The common garden pea (Pisum sativum L., Fabaceae) has a morphologically complex leaf that is odd pinnately compound and possesses three different types of lateral appendages: stipules, leaflets, and tendrils. The typical adult pea leaf blade is dissected into two to three pairs of proximal leaflets, three pairs of distal, simple tendrils, and a terminal, simple tendril (Fig. 1A). The gene known to have the greatest effect on the compound nature of pea leaf development is not a KNOX1 gene but is UNIFOLIATA (UNI), which is a member of the LEAFY/FLORICAULA gene family. The UNI gene is known to be upregulated by both auxin and GA (Bai and DeMason, 2006; DeMason and Chetty, 2011). In fact, the simpler nature of the leaves of GA mutants of pea demonstrate a positive correlation between GA levels and leaf complexity, and GA application increases the number of pinna pairs produced on leaves of the GA mutants and on leaves of a mild uni mutant, known as uni-tac (DeMason, 2005a; DeMason and Chetty, 2011). Since the KNOX1-ARP module seems to be conserved in plant species, one would expect to see similar leaf phenotypes in KNOX1 overexpression and ARP loss of function mutants in different species. Although there are some weak consistencies, the differences have “led to uncertainty regarding their (the ARP genes) precise role in leaf development” (Piazza, Jasinski and Tsiantis, 2005, p. 698 ). The weak consistencies in leaf phenotypes seen in KNOX1 overexpression/ARP loss of function mutants include: loss or partial loss of ability to produce a lamina (i.e., unifacial leaves), especially the displacement of the lamina to the distal region of the leaf producing “trumpetshaped” or peltate leaves in Antirrhinum, Lycopersicum, and Nicotiana (Waites and Hudson, 1995; Kim et al., 2003; McHale and Koning, 2004), and production of epiphyllous structures or “knots” on leaf blades in Zea mays, Antirrhinum, Arabidopsis, and Nicotiana (Freeling and Hake, 1985; Waites and Hudson, 1995; Byrne et al., 2000; McHale and Koning, 2004). Anatomical preparations of “knots” have revealed that they do not have similar histologies in different species. In Zea, “knots” consist of sheath/auricle or ligule tissues ectopically positioned on the blade (Hake, 1992; Ramirez et al., 2009). In Antirrhinum, “knots” consist of regions in which adaxial cell types are replaced by abaxial ones (Waites and Hudson, 1995), and in Nicotiana, “knots” consist of ectopic, miniature, bifacial leaf blades produced along the midrib (McHale and Koning, 2004). “Knot” anatomy in maize represents another example of proximal structure displacement into distal regions of the leaf. Pisum is an important species to study divergent regulatory pathways of compound leaf development since GA has a different effect, and the pea KNOX1 genes are not the primary regulators of pinna initiation. A significant unresolved question is whether there is interaction between the KNOX1-ARP module and UNI functions in pea leaf morphogenesis. Although there are no known pea KNOX1 mutants, there are mutant alleles in the pea ARP gene known as CRISPA (CRI), which can be used

409

to start addressing this question. The first crispa (cri-1) mutant was named by Lamm (1949) because the leaflets were “characterized by their crisp or folded” form. Leaflets fold abaxially due to linear epiphyllous extensions, or “knots”, on the adaxial surface resulting in pleats. The histology of these “knots” has not been studied. Tattersall et al. (2005) identified CRISPA as the pea ARP gene, found two additional mutant alleles (cri-2, cri-3), described the phenotypes and identified a KNAT1/BPlike KNOX1 gene, PsKN2, as its antagonist. Later, Sainsbury et al. (2006) identified a fourth mutant allele (cri-4) that they concluded was a null mutant. We had grown cri-1 plants in the greenhouse to observe their phenotype and felt many aspects were misinterpreted by Tattersall et al. (2005) and warranted re-examination. Understanding the developmental role that genes play is predicated on accurate morphological and anatomical interpretations of mutant phenotypes. Comparisons must be done under more tightly controlled conditions and with superior anatomical techniques than were previously used. To accomplish this, we controlled for genetic background effects by making a set of 6 new isogenic lines that included cri-1, controlled for the heteroblastic effects by using an equivalent, adult reference leaf for all genotypic comparisons, and collected leaf tissue samples in comparable regions of these leaves before using aldehyde fixation and resin preparations for anatomical evaluation. Further, we extended the study by comparing expression of some important genes, known to affect pea leaf morphogenesis, in cri-1 and wildtype shoot tips. In our study, we specifically addressed the following questions: What are the effects of the crispa (cri) mutation on the morphology of pea leaves, especially on the three planes of leaf asymmetry? What are the morphological and anatomical manifestations of “knots” on leaves of the cri mutant? What roles do GA, HOP1 and UNI play in development of the cri leaf phenotype? MATERIALS AND METHODS Plant materials—The Marx (1974) leaf architecture isogenic lines were extended by introducing crispa into them. The recurrent parents were wildtype (WT)–W6 22593, tendrilless (tl)–W6 22594, afila (af)–W6 22597, tendrilled acacia (uni-tac)–W6 27606, and stipules reduced (st)–W6 22595. The donor parent was W6 15292 (CRI/−). This latter line, which possessed the crispa-1 (cri-1) allele (Lamm, 1949), was also obtained from the Marx collection (USDAARS Pacific West Pisum germplasm collection, http://www.ars.usda.gov/Main/ docs.htm?docid=15144). Backcross breeding procedures—CRI/− (W6 15292) was grown, and seed was collected only from plants with the cri phenotype to use as the donor parent. The donor parent was backcrossed to the recurrent parents. Twenty pollinations were done for each combination, and the resulting F1 seeds were pooled. F2 plants with the cri phenotype were selected from each backcross before repeating the pollinations. Six backcrosses were made in total for each combination. The cri lines were used as the female parent for five of these backcrosses and as the male parent for one backcross for each combination to remove any maternal effects of the original donor parent. Finally, cri-1 tl was crossed with af cri-1, and triple mutants were selected in the F2 generation. BC6F3 seeds from the final backcross for each new genotype were propagated. The five newly created lines were deposited into the Marx collection mentioned above and are W6 46469 (cri-1), W6 46470 (cri-1 uni-tac), W6 46471 (cri-1 af), W6 46472 (cri-1 tl), W6 46473 (cri-1 af tl), and W6 46474 (cri-1 st). Construction of DR5::GUS cri line—WT DR5::GUS lines were constructed using WT (W6 22593) for transformation as described previously (DeMason and Polowick, 2009). The RTP9 reporter insert line was then crossed into the new cri isogenic line (above) as described previously (DeMason and Chetty, 2011). Selection over several generations produced a DR5::GUS cri line, which was homozygous for the insert as well as for cri-1.

410

AMERICAN JOURNAL OF BOTANY

[Vol. 101

Fig. 1. Comparisons of (A–F) leaves, (B) shoots, and (G) flowers of (A, G) Wildtype (WT) and (B–G) the crispa (cri) mutant. Asterisk (*) indicates ectopic stipule; arrowheads indicate petiole–rachis curvature; arrow indicates leaflet tip modification; dl = doubled leaflet, k = “knots”; s = stipule.

March 2014]


Greenhouse and growth chamber conditions—For standard propagation and phenotypic analysis, two plants were grown per 1-gal pot with UC soil mix supplemented with Osmocote 16-16-16 (Scott, Marysville, Ohio, USA). The plants were grown in natural sunlight in a greenhouse constructed of clear glass with temperatures maintained between 15° and 32°C. An automatic watering system provided soil saturation every other day. For phenotypic characterization, seeds were sown in late August. For gene expression analysis, plants were grown in an EGC model GC15 growth chamber (Chagrin, Ohio, USA), with a combination of incandescent, halogen, and sodium light bulbs, with 8 h light/16 h dark, and 20°C day and 15°C night regime as described previously (DeMason and Chetty, 2011; DeMason et al., 2013). Photosynthetically active radiation (PAR) was approximately 340 µmol·m−2·s−2 at plant level. Leaf form comparisons and measurements—Sixteen to 18 plants were grown of each of the following genotypes: WT, st, cri, af cri, af cri tl, cri st, cri tl, and cri uni-tac. All plants with cri genotypes were observed weekly, and when all plants were in flower, detailed notes were made on the fully expanded (i.e., mature) last vegetative leaf or that at the first flowering node. Henceforth, these leaves are designated adult reference leaves or reference leaves. Observations on reference leaves included the number and types of all pinnae and their characteristics, relative leaf curvature, characteristics of stipules, locations of any ectopic stipules, presence of any leaflet abnormalities and their positions, and presence of “knots.” These reference leaves were also used for whole leaf photographs, leaf parameter measurements, and collection of tissue samples for anatomical analysis. Petiole length, height (adaxial surface to abaxial surface), and width (lateral surface to lateral surface) were measured on reference leaves with a millimeter ruler on WT, st, and cri st plants that were fruiting (N = 18). Height and width measurements were made in the middle of the petiole. Petiole length and height/ width means were subjected separately to a 1-way ANOVA. Since both results showed significant differences (P ≤ 0.0001), the posthoc Student-NewmanKeuls multiple comparison test was run to compare each pair of parameters within each data set using GraphPad Prism (GraphPad Software, La Jolla, California, USA). Significance was defined at the 0.05% level. Leaflet lengths and widths were measured with a millimeter ruler on a leaflet at the lowermost pinna position of mature, reference leaves on WT and cri plants (N = 16). Pinna widths were measured across the widest part of the leaflet. Doubled leaflets on cri leaves were not used. Student t tests were performed using GraphPad Prism, and significance was defined at the 0.05% level. Photography of whole plants, leaves, and leaflets—Whole plants were photographed in the greenhouse with natural sunlight. A piece of black, unwaled corduroy was taped to a rectangular polyvinylchloride (PVC) pipe frame as a background and photographs were taken with a Nikon Coolpix L110 camera affixed to a standard tripod. Fully mature reference leaves were detached from greenhouse-grown plants and placed in plastic bags containing wet paper towels during transport to the laboratory. Leaves were photographed using a Canon Powershot G3 camera mounted on a CS-5 Testrite (Newark, New Jersey, USA) copy-stand on black velveteen in an Aristo DA-17 (Aristo Lamp Products, Port Washington, NY) light box. Leaflets possessing distinct “knots” on plants grown in the growth chamber were removed from 1-mo.-old seedlings transported to the laboratory in flats and photographed on a Nikon SMZ-10 dissecting microscope with reflective light from a fiber optic light source and using an Infinity 1 digital camera.

411

were placed immediately in buffered aldehyde fixative (4% formaldehyde and 2.5% glutaraldehyde v/v in 50 mmol/L phosphate buffer, pH 7.2) and then placed in a refrigerator for 2–3 d. Samples for fixation were selected in the center of the leaf part used, except for leaflets where regions specifically possessing “knots” were collected. Fixed specimens were subsequently rinsed in buffer, dehydrated through a graded ethanol series and embedded in JB-4 or JB-4 Plus methacrylate (Polysciences, Warrington, Pennsylvania, USA). Sections were cut on a Sorvall JB-4A microtome at 4 µm thick, stained with 0.05% toluidine blue (C.I. 52040) in 50 mmol/L benzoate buffer pH 4.4, and mounted with coverslips using Permount. Observations and photographs were made on a Nikon Multiphot macrophotography unit using transmitted light or on either a Zeiss Standard or Zeiss Universal bright-field microscope with an Infinity 1 digital camera. Digital images were edited and assembled into plates using Adobe (San Jose, California, USA) Photoshop CS. GUS activity analyses—Plants were grown in the growth chamber until leaves at nodes 10–12 were exposed (approximately 1 mo) as described previously (DeMason and Chetty, 2011; DeMason et al., 2013). Shoot tips, leaf primordia, and individual leaflets were stained and prepared as described previously (DeMason and Polowick, 2009). Samples were observed on a Nikon SMZ-10 dissecting microscope with reflective light from a fiber optic light source and photographs were taken with an Infinity 1 digital camera. Quantitative assays of shoot tips were done as previously described (DeMason and Polowick, 2009; DeMason et al., 2013.) A Student t test was performed to compare the means using GraphPad Prism, and significance was defined at the 0.05% level. qRT-PCR of gene expression—Six genotypes from the set of near-isogenic lines (WT, af, tl, af tl, cri, and uni-tac) were used for analysis. Plants were grown in the growth chamber and shoot tips were collected when leaves at nodes 10–12 were exposed as described previously (DeMason et al., 2013). Total RNA was isolated, then first strand cDNA synthesis and SYBR green assays were carried out as described previously (DeMason and Chetty, 2011). The cDNA was diluted 1 : 100, and 1.0 µL of the dilution was used in a SYBR green quantitative RT (qRT)-PCR. The primers used for the genes investigated are listed in Appendix S1. The cycling conditions were as described in DeMason and Chetty (2011) and DeMason et al. (2013). After PCR amplification, melting curves were generated as described previously (DeMason and Chetty, 2011; DeMason et al., 2013). Expression levels for UNI, TL, and five other genes known to be expressed in pea shoot tips were determined for WT and cri-1 shoot tip comparisons. Homeodomain Protein1 (HOP1) was identified as the STM-like gene in pea (Giles et al., 1998) and mapped to linkage group IV (Weeden and DeMason, 1999); however, it was later identified, remapped to the same linkage group and renamed PsKN1 by Hofer et al. (2001). Mendel’s tall plant gene (LE), the auxin exporter gene (PIN1), and TL are auxin-regulated genes. Expression of the NO APICAL MERISTEM1/2 (NAM1/2) gene and the CUP-SHAPED COTYLEDON3 (CUC3) gene was also quantitated. Expression levels for CRI and HOP1/PsKN1 were compared in WT and af, tl, af tl, st, and uni-tac. ACTIN was used as the reference gene. For statistical analysis, ΔΔCt was calculated comparing the control to the mutant genotypes, and P values for Student’s t tests were calculated using Microsoft (Redmond, Washington, USA) Excel as described in Yuan et al. (2006).

RESULTS Leaflet clearings—Leaflets from the proximal-most position of mature, adult reference leaves of WT and cri greenhouse-grown plants were removed and fixed in FAA (4% formaldehyde, 5% acetic acid and 70% ethanol, v/v) in the refrigerator for 2–3 d and then maintained in fixative at room temperature until most of the chlorophyll had been extracted. Samples were rinsed in 70% ethanol and allowed to clear further if needed, then processed through a graded ethanol series to 100% ethanol and stained overnight in a saturated solution of safranin 0 (C.I. 50240) in absolute ethanol. Stained samples were quickly passed through a graded ethanol– xylene series to 100% xylene. Leaflets without “knots” were mounted on microscope slides with coverslips and photographed on either a Nikon Multiphot macrophotography unit with transmitted light or on a Zeiss Universal light microscope with bright-field optics and an Infinity 1 digital camera. Leaflets with “knots” were stored in absolute xylene and photographed on a Nikon SMZ-10 dissecting microscope with reflective light from a fiber optic light source and photographs were taken with an Infinity 1 digital camera. Leaf anatomy—Leaf parts (petioles, stipules, leaflets, and ectopic stipules) were dissected from adult reference leaves on greenhouse-grown plants. Samples

Leaf and plant morphology of WT vs. cri— Mature, adult leaves on WT plants are odd, pinnately compound and possess three types of lateral appendages: stipules at the leaf base, leaflets, and tendrils in the blade. The leaf blade typically possesses two to three pairs of proximal leaflets, three pairs of distal, simple tendrils, and a terminal, simple tendril (Fig. 1A). An obvious morphological abnormality of reference leaves on cri plants was the extreme adaxial curvature of the shortened petiole and proximal rachis, which sometimes exceeded 180° (Fig. 1B arrowheads, 1C, F). True stipules were fused to the petiole along its entire length (Fig. 1C, E, F), and ectopic stipules occurred at the base of all leaflets, the most proximal pair was also fused along the distal margin of the true stipule (Fig. 1E). The presence of ectopic stipules commonly occurred beneath the proximal-most pair of tendrils as well (47% of reference leaves).

412


Leaflets on cri leaves were narrower than those of WT, the margins were abaxially recurved, and they possessed “knots” on the adaxial surface (Fig. 1D arrows). Leaflets at the proximalmost position most commonly possessed “knots”. One or both of the basal leaflets were often doubled (i.e., two leaflets attached to the same position) (44% of reference leaves). Typically, the leaflets attached at the same position were completely or partially fused (Fig. 1B, C, 9F). The proximal-most leaflet of each doubled pair was always the smaller, and often it was folded under at the fusion zone. The leaflets at the proximal-most position sometimes had V-shaped notches in their distal tips and a short, central tendril similar to the insecatus mutant of pea (Berdnikov et al., 2000) (25% of reference leaves) (Fig. 1C, arrow). This leaflet feature is similar to bifurcated leaves observed on the Kn1-DL mutants of maize (Ramirez et al., 2009). Finally, a few leaflets had adaxially thickened bases and somewhat “trumpet-shaped” blades. However, the numbers and basic types of pinnae (leaflets vs. tendrils) were normal and typically consisted of two pairs of proximal leaflets, three pairs of distal, simple tendrils, and terminal tendril. The mean number of pinna pairs on reference leaves of cri plants was 5.2 (N = 16), which was similar to the estimated 5–5.5 for the WT line from previous publications (DeMason and Villani, 2001; DeMason, 2005b). The form of distal rachis segments and simple tendrils were not obviously different from those of WT leaves. Early juvenile leaves and leaves subtending axillary inflorescences on cri plants were distinctly less abnormal or stunted than the adult reference leaves. The inflorescences and flowers on cri plants were typically indistinguishable from those of WT with the exception of cri uni-tac double mutants where shape and orientation of wing and keel petals were often abnormal resulting in exposure of keel petals and poor seed set (Fig. 1G). Leaf and plant morphology of stipules reduced vs. cri stipules reduced double mutants—Leaf form of the stipules reduced (st) mutant has not been studied extensively. A pair of stipules was present on every leaf, but reduced in size (Fig. 2A, B). Often the stipules were spatulate in shape with miniaturized “stipules” at their bases (Fig. 2B). Unlike stipules on WT plants, they were bilaterally symmetrical rather than bilaterally asymmetrical. Double mutants of st and cri were created to test for interactions between the genes. Leaves on cri st plants possessed very narrow, linear stipules, which were not fused to petioles (Fig. 2C–E). Petioles on these double mutants were not abaxially curved as they were on cri plants, and no ectopic stipules were present on the rachis proximal to leaflets or tendrils (Fig. 2D, E). Leaflets were narrow, margins were abaxially recurved, and “knots” were present on the adaxial surface (Fig. 2D, E). One or both of the basal leaflets on cri st plants were often doubled (88% of reference leaves) (Fig. 2D). One reference leaf (of 17) had two types of pinnae (compound tendril and leaflet) attached to the same basal position (Fig. 2E). A notched tip with a short central tendril was common on basal leaflets of this genotype (41% of reference leaves) (not shown). One reference leaf (of 17) had a trumpet-shaped leaflet. The numbers and basic types of pinnae were normal, and most adult leaves had two pairs of leaflets, three pairs of simple tendrils, and a terminal tendril. The mean number of pinna pairs per reference leaf was 5.0 (N = 17). The form of distal rachis segments and simple tendrils were not obviously different from those of st leaves. Effects of cri and uni-tac, tl, af, and af tl on leaf form— The cri allele was crossed into additional pea leaf form mutants to test for interactions between genes. Reference leaves on cri uni-tac

[Vol. 101

plants had adaxial curvature of the petiole–proximal-rachis axis, stipules fused to shortened petioles, narrow leaflets with abaxially recurved margins, “knots”, and ectopic stipules (Fig. 3A, B). Most reference leaf blades possessed two lateral pinna pairs (mean = 2.05; N = 18) and a terminal leaflet. This mean was lower than the estimated means of 2.8 and 3.5 for leaves at this nodal position on uni-tac plants from previous studies (DeMason and Schmidt, 2001; DeMason, 2005b). A pair of leaflets always occurred in the basal (proximal) lateral, pinna position, but the upper (distal) lateral, pinna position possessed a pair of leaflets (Fig. 3A, left), a pair of simple tendrils (Fig. 3A, right, 3B), or a mixed pair (leaflet opposite a tendril) (Fig. 3A, center). The size of the rachis segment between the two pinna pairs on reference leaves of cri uni-tac correlated with pinna form at the distal, lateral position: it was short and wide between leaflet pairs, long and narrow between leaflet and tendril pairs, and intermediate between leaflet and mixed pairs (Fig. 3A, B). Ectopic stipules occurred beneath lateral leaflets, but typically, not beneath lateral tendrils. “Knots” occurred on lateral as well as terminal leaflets. Basal leaflets were never doubled, but the terminal leaflets often were lobed due to pinna fusion, which is typical of uni-tac mutants (DeMason and Schmidt, 2001). This was the only genotype in this set of isogenic lines that had consistently deformed wing and keel petals (Fig. 1G), and the number of pods produced was severely reduced on a per plant basis. Reference leaves on cri tl plants had adaxial curvature of the petiole–proximal rachis axis, stipules fused to shortened petioles, narrow leaflets throughout with only slightly recurved margins, and ectopic stipules (Fig. 3C, F). “Knots” were rarely present, and when present, occurred on the basal-most leaflets (Fig. 3C, leaf on left). Reference leaf blades possessed 5–6 (mean = 5.56; N = 16) pairs of lateral leaflets and a terminal leaflet. The mean number of leaflets pairs previously observed on the tl genotype was 5.2 (DeMason, 2005b). On leaves with five leaflet pairs, ectopic stipules occurred beneath the basal two or three leaflet pairs. On leaves with six leaflet pairs, they occurred beneath the basal three or four leaflet pairs. The rachis segments between leaflets with ectopic stipules were shorter than the rachis segments between leaflet pairs lacking them. Basal leaflets were doubled on about 50% (8/16) of all reference leaves, and 50% (4/8) of these doubled leaflets were unfused. Many different abnormal leaflet forms occurred on reference leaves of this genotype, and they were typically leaflets in the distal region of the leaf (distal, lateral pairs, or terminal leaflet). These included leaflets with extended radial tips (Fig. 3C, leaf on left, 3E), trumpet-shaped leaflets (Fig. 3D), fusions between leaflets (Fig. 3G), leaflets reduced to short bristles, and missing leaflets (Fig. 3E). Reference leaves on af cri plants had extreme adaxial curvature of the petiole–proximal rachis axis (sometimes 360° or more), stipules fused to shortened petioles, ectopic stipules, and very stunted leaves (Fig. 3H). Some adult reference leaves never expanded to any extent before senescing and turning yellow (Fig. 3H, leaf on right). Reference leaves had compound tendrils in the proximal blade region and simple tendrils in the distal like those of the af genotype, and the mean number of pinna pairs was 5.0 (N = 16). This mean is close to the estimated mean number of pinna pairs on the af line, of 5.1 determined in a previous study (Villani and DeMason, 1999a). Ectopic stipules occurred beneath the proximal compound tendrils, but not beneath the distal simple tendrils. Fusions between ectopic stipules and the basal-most portion of compound tendrils occurred commonly. No “knots” were produced on leaves of this genotype. There was so much stunting and lack of elongation in the proximal region of leaves that it was not possible to determine whether basal pinnae were doubled.

March 2014]


413

Fig. 2. Comparisons of (A, C) shoots, (B) stipules, and (A, B) leaves of reduced stipule (st) and (C–E) cri st double mutants. Asterisk (*) indicates basal “stipular” flap; ct = compound tendril; dl = doubled leaflet; dp = doubled pinna (compound tendril and leaflet); i = internode; inf = axillary infructescence; k = “knots”; l = leaflet; p = petiole; s = stipule.

414


[Vol. 101

Fig. 3. Leaf phenotypes of (A, B) cri uni-tac double mutants, (C–G) cri tl double mutants, (H) af cri double mutants, and (I) af cri tl triple mutants. Asterisk (*) indicates ectopic stipules; arrows indicate leaflet modifications; dl = doubled leaflet; k = “knots”; p = petiole; s = stipule; tl = terminal leaflet. Scale bars = 4 cm. Scale bar in D represents scale in D–G.

March 2014]


Reference leaves on af cri tl plants had extreme adaxial curvature of the petiole–proximal rachis axis (sometimes 360° or more), stipules fused to shortened petioles, very narrow leaflets with recurved margins, ectopic stipules, and extremely stunted leaves (Fig. 3I). Many reference leaves failed to expand to any extent (Fig. 3I, center leaf) so that morphological evaluation was not possible. Only the distal half of some leaves expanded, and it was clear that the pinnae were compound with miniaturized, terminal leaflets. The leaflets on this genotype had very little laminar expansion and a toothed tip; many were needle-like but with no discernible thigmotropic response; or were lacking entirely (Fig. 3I, leaf on right). No “knots” were produced on leaflets of this genotype. Petiole and leaflet size comparisons—Because stipule–petiole fusion did not occur on leaves of the cri st double mutant, we could use this genotype to determine the effect of the cri mutation on petiole development. Petiole length of reference leaves was reduced on st leaves compared to WT and further reduced on leaves of the double mutant (Fig. 4A). Petiole heights and widths did not

Fig. 4. (A) Petiole and (B) leaflet dimension comparisons. Means and standard deviations of 18 replicate samples are shown in (A). Petiole lengths differed significantly as determined by ANOVA followed by Student– Newman–Keuls multiple comparison test (F2,51 = 35.94; P < 0.0001), and heights and widths with different letters differed significantly as determined by the same tests (F5,102 = 24.26; P < 0.0001). Means and standard deviations of 16 replicate samples are shown in (B). Means of leaflet lengths compared using a Student t test were found to be similar (t = 1.532; df = 30; P = 0.1361). Means of leaflet widths were also compared and found to be significantly different (t = 9.371; df = 30; P < 0.0001).

415

differ within or between the WT and st genotypes, but cri st petioles were distinctly taller in the adaxial–abaxial direction compared with their widths and were also significantly taller than petioles of the other two genotypes in the comparisons (Fig. 4A). Leaflets on reference leaves of cri and WT plants were similar in length, but cri leaflets were distinctly reduced in width (Fig. 4B). The length to width ratios of WT and cri leaflets was 1.60 and 2.57, respectively, (or width to length ratios were 0.62 and 0.39, respectively, to compare with data presented inby Tattersall et al. [2005]). Stipule morphology and anatomy—The arrangements of cell types in transverse sections of stipules of reference leaves on WT plants was distinctly bifacial according to the standard anatomical definition (Cutter, 1971; Esau, 1977). A single layer of palisade parenchyma occurred on the adaxial side, and approximately four layers of spongy mesophyll occurred on the abaxial side of the ground tissue system (Fig. 5A, B). The median vascular bundle as well as minor ones had xylem on the adaxial side and phloem on the abaxial side, and fiber bundles flanked each (Fig. 5A). Bundle sheath cells surrounded all veins. Small vascular bundles in the lamina were positioned at the interface between the palisade and spongy mesophyll (Fig. 5B). Stomata occurred in both the adaxial and abaxial epidermises. Stipules on reference leaves of the st genotype also had bifacial anatomy. All cell types appeared to be smaller than those in WT, but the same number of cell layers were typically present (Fig. 5C, D). Cell types, sizes, and arrangements were very similar in cri stipules compared with WT; therefore, they were also distinctly bifacial (Fig. 5E, F). Stipules on this genotype tended to have an undulate adaxial surface because it “bulged” above large vascular bundles (Fig. 5F). The much-reduced stipules on reference leaves of the cri st double mutant had bifacial tissue arrangement in the vascular bundles (Fig. 5G). However, since the spongy mesophyll cells were fairly uniformly spherical in shape and the palisade parenchyma cells were shorter and wider than those of cri stipules, these two cell types in the ground tissue system were inconsistently distinguishable (Fig. 5G, H). The anatomy of ectopic stipules differed from that of true stipules in that there were no median bundles, no palisade parenchyma, and the mesophyll cells were smaller than those of true stipules. The small vascular bundles had a bifacial arrangement of xylem and phloem, and the surface was undulate like true stipules on cri leaves (Fig. 5I, J). The lack of bifacial histology in the ground tissue system was more strongly correlated with stipule size than with the presence of the cri allele. Petiole morphology and anatomy— The petiole of WT reference leaves was relatively circular in outline and was hollow (Fig. 6A). Bifaciality was subtly manifested in several ways: (1) the adaxial surface was slightly flattened, (2) the abaxial surface was slightly ribbed due to the presence of collenchyma below the median abaxial bundle, and (3) the distribution of the five major vascular bundles was two on the adaxial, two lateral, and one on the abaxial sides (Fig. 6A–C). Minor vascular bundles occurred between the five major bundles, including on the adaxial side of the petiole. The xylem of each bundle faced the center, and the phloem faced the epidermis. The petioles were green, and chlorenchyma occurred all the way around the petiole immediately under the epidermis (Fig. 6B, C). Petiole form of the af and cri genotypes differed from WT in that there were distinct grooves on their adaxial surfaces. The much larger petiole of the af genotypes had more major bundles than the petioles of WT and the groove was associated with a median adaxial bundle (Fig. 6D). Like WT petioles, the xylem faced the center,

416


[Vol. 101

Fig. 5. (A, C, E, G) Stipule anatomy including a major or median vein and (B, D, F, H, I, J) central lamina (A, B) of WT, (C, D) st mutant, (E, F) cri mutant, (G, H) cri st double mutant, and (I, J) cri ectopic stipule. f = bundles of fibers, m = generalized mesophyll, ph = phloem, pp = palisade parenchyma, sm = spongy mesophyll, x = xylem. Scale bars = 250 µm.

March 2014]


417

Fig. 6. Petiole anatomy of (A–C) WT, (D, G) af mutant, (E, F, H) cri mutant, and (I–L) cri st double mutant. Scale bar in A also represents scale in E, F, I, J; in B also represents C; in K also represents L. Asterisks (*) indicate major vascular bundles, c = collenchyma; arrows point to callus growth (J) or evidence of recent adaxial cell division activity (K).

418


and there was chlorenchyma all around the perimeter of the petiole (Fig. 6G). Petioles of adult cri leaves were also hollow, but were much taller than wide and were typically contorted due to curvature (Fig. 6E). There were five major vascular bundles, two pairs of lateral bundles, and one median abaxial bundle, and the stipule was fused to the grooved adaxial surface. The upper lateral bundles were distinctly larger than the median abaxial bundle, unlike petioles of WT or af. The xylem side of all petiole bundles faced the center, and chlorenchyma and minor bundles were lacking on the adaxial side (Fig. 6F, H). Anatomy of the stipule adjacent to the petiole consisted of vascular bundles with xylem facing the adaxial surface, and the ground tissue lacked palisade parenchyma cells (Fig. 6F, H). A few millimeters away from the petiole/stipule fusion zone, stipules of cri leaves possessed an adaxial palisade layer in the ground tissue and were bifacial as described above (data not shown). Petioles on reference cri st leaves were hollow and possessed a ring of vascular bundles in which xylem faced the center (Fig. 6I–L). Like WT, there were five major bundles: two adaxial, two lateral, and one median abaxial bundle. The two adaxial bundles were often greatly enlarged, and in one specimen, the phloem almost surrounded the xylem (not shown). The adaxial surface was grooved and possessed minor vascular bundles, unlike cri petioles. A unique feature of the petioles of this genotype was that calluslike “knots” were present on the adaxial surface in many sections. There was much variation in the histology of the adaxial cells in different petioles of this genotype, which included extensive adaxial cell division activity producing callus (Fig. 6J) to limited activity that resulted in rupture of the epidermis (Fig. 6K) to relatively little ectopic cell division activity (Fig. 6L). None of the cri st petioles had chlorenchyma on their adaxial side. Leaflet morphology and anatomy— Leaflets on reference leaves of WT plants were traversed by many orders of veins. Secondary veins that diverged from the midrib, some of which ultimately terminated in marginal teeth were the highest order. Lower orders of veins surrounded areoles which also contained freely ending veinlets (Fig. 7A, C). In WT leaflets, secondary veins diverged at a large acute angle from the midrib, areoles were broad, and marginal teeth were widely spaced (Fig. 7A, C). In contrast, in cri leaflets, secondary veins diverged at a small acute angle from the midrib, areoles were narrow and poorly expanded, and marginal teeth were closely spaced (Fig. 7B, D). WT leaflets possessed loops of minor veins that ran along the margins including the edges of obtuse marginal teeth and in the sinuses between teeth, which may provide structural reinforcement of the edge (Fig. 7A, C). The cri leaflets generally lacked this marginal reinforcement, which may have contributed to the abaxial curling of leaflet margins described above (Fig. 7B, D). Leaflet clearings illustrated the fact that “knots” on the adaxial surface of cri leaflets had vascularization, which was often similar to that of marginal teeth (Fig. 7E–G). The anatomy of WT leaflets has been described previously, and our observations were consistent with past descriptions (Villani and DeMason, 1997, 1999b). The adaxial and abaxial epidermises had stomata; the mesophyll was characterized by the presence of palisade parenchyma on the adaxial side and spongy mesophyll on the abaxial side; and the vascular bundles consisted of xylem on the adaxial side and phloem on the abaxial (Fig. 8A). Bundle sheath extensions commonly abutted both epidermises on larger bundles. Like stipules, small vascular bundles in the lamina of leaflets were positioned at the interface between the palisade and spongy mesophyll (data not shown). Leaflet anatomy of cri reference leaves was distinctly bifacial since palisade and spongy mesophyll, and xylem

[Vol. 101

and phloem were present in the appropriate positions (Fig. 8B). The histological differences between WT and cri leaflets were that the palisade cells were shorter in cri than in WT, and the leaflet surface of cri was slightly undulate due to the presence of mesophyll cells above major vascular bundles (Fig. 8B). Mesophyll cell widths appeared similar in cri and WT leaflets. “Knots” were commonly present in the proximal region of leaflets on cri plants: in the regions between secondary veins (Fig. 9A, B, D, E or between the outer most secondary vein and the margin (Fig. 9C, D) and were particularly characteristic of the fusion zone between doubled leaflets (Fig. 9F). “Knot” morphology was variable and consisted of rows of narrow, parallel ridges (Fig. 9B), ectopic toothed margins, which were common (Fig. 9A, C), elevated ridges or platforms (Fig. 9B, E), and ectopic leaflets (Fig. 9D, F). Most “knots” had adaxial coloration (dark green) (Fig. 9A, B, C, D), but some had abaxial coloration (light green) (Fig. 9B, D, F). The presence of palisade parenchyma on the adaxial surface correlated with adaxial coloration. Regions of narrow, parallel ridges had an undulating adaxial surface with distinct palisade parenchyma in the ridges (Fig. 9G, K, right side), and this was also true of some more elevated ridges and larger extensions (Fig. 9J). The palisade cells in these “knots” were continuous with the palisade cells in adjacent unaffected areas of the leaflet. However, some elevated ridges had spongy parenchyma on the uppermost surface and palisade parenchyma on the lateral or lower surfaces (Fig. 9K, left side). Additional vascularization was associated with the elevated regions of all “knots”, and the added smaller vascular bundles were positioned at the interface between palisade and spongy mesophyll (Fig. 9G, K, arrows). Ectopic, toothed, marginlike extensions typically had palisade parenchyma all the way around the edges (Fig. 9H) and more extensive internal vascularization. Finally, ectopic leaflets had bifacial anatomy that was reversed (mirrored image) of the leaflet itself (Fig. 9I, K). Vascular bundle anatomy in these “knots” was also reversed with xylem facing the palisade layer (Fig. 9I insert). The peripheral palisade layers of these “inverted knots” were also continuous with the palisade layer of the unaffected leaflet (Fig. 9I, K). GUS expression from DR5::GUS—DR5::GUS is a reporter gene that provides quantitative measurements of auxin response and allows localization of high auxin response. Patterns of GUS staining in DR5::GUS cri shoot tips, and early leaf development were similar to WT as previously described (DeMason and Polowick, 2009; DeMason et al., 2013) (data not shown). GUS activity in shoot tips of WT (0.455 nmol·mg−1·min−1 ± 0.022) and cri-1 (0.394 nmol·mg−1·min−1 ± 0.015) was similar (F = 2.23; P = 0.086). During late stages of leaf development when leaflets were expanding, GUS staining was associated with marginal teeth (Fig. 10A) and procambial strands around developing areoles (Fig. 10A–C) as reported previously (DeMason and Polowick, 2009; DeMason et al., 2013) and in the tips of maturing “knots” and in procambial strands developing within them on the adaxial surface of leaflets (Fig. 10A–E). GUS staining was associated with all “knot” manifestations: in tooth-like regions of ectopic margins (Fig. 10A, C, D) and in the tips of isolated projections (Fig. 10B, E). GUS staining remained longer within the tips of “knots” than it did in differentiating procambial strands (Fig. 10D, E). Gene expression comparisons between WT and cri shoot tips—Expression levels of four genes (NAM1/2, CUC3, TL, and PIN1) known to be upregulated by auxin were reduced in cri compared to WT shoot tips (Fig. 11). The mRNA level of LE, which is known to be downregulated by GA and upregulated by auxin, was

March 2014]


419

Fig. 7. Leaflet clearings of (A, C) WT and (B, D–G) the cri mutant. Black arrows indicate “knots”. Scale bar in A also represents B; in C also represents D; and in E also represents F, G.

420


[Vol. 101

known to play roles in leaf development of many plant species, including the lycopod Selaginella (Harrison et al., 2005). The module’s presence in pea prompts further questions regarding its role in pea leaf morphogenesis and whether it intersects the UNI developmental pathway. These questions motivated our detailed reinvestigation into the phenotype of cri mutants of pea to specifically determine (1) the role that the CRI gene plays in pea leaf morphogenesis, especially in regulating the three planes of leaf asymmetry, (2) the morphological and anatomical manifestations of “knots” on leaves of the cri mutant, and (3) which gene and hormone interactions contribute to the cri leaf phenotype.

Fig. 8. Leaflet anatomy of (A) WT and (B) cri plants. f = bundles of fibers, ph = phloem, pp = palisade parenchyma, sm = spongy mesophyll, x = xylem. Scale bars = 100 µm.

also reduced in cri compared to WT (a two-thirds reduction). Two genes had higher mRNA levels in cri shoot tips compared with WT: HOP1/PsKN1 (almost a 2-fold increase) and UNI (a 3-fold increase). CRI expression was somewhat reduced in tl, af, and af tl shoot tips compared with WT, but it was very much reduced in st and uni-tac (Fig. 12A). HOP1PsKN1 expression was elevated in tl, similar in af tl, and reduced in af, st, and unitac shoot tips compared with WT (Fig. 12B). DISCUSSION Compound leaves are thought to have evolved independently multiple times within the angiosperms (Bharathan and Sinha, 2001; Doyle, 2007; Sanders et al., 2009). The developmental challenge is to determine whether similar or different developmental processes have been recruited to produce similar morphologies in different evolutionary lines. Pea provides a unique opportunity to study a divergent mechanism of developing compound leaves in comparison to the tomato model because (1) it has acropetal pinna initiation compared with basipetal; (2) it uses the LFY/FLO family gene, UNI, instead of KNOX1 to regulate leaf complexity; and (3) GA promotes increased complexity rather than suppressing it. In spite of these developmental differences, pea possesses the conserved KNOX1/ARP module

The CRI gene and the three planes of leaf asymmetry in pea— Pea leaves are an extreme example of proximal–distal asymmetry because they possess stipules and two forms of pinnae in the blade (bifacial leaflets in the proximal region and unifacial tendrils in the distal) and their size/weights decrease acropetally (Lu et al., 1996; Villani and DeMason, 1997, 1999a, b, 2000; DeMason and Schmidt, 2001). This morphology is created by gradients of hormones (auxin and GA) and gene expression (UNI, AF, TL) (DeMason and Chawla, 2004a, b; DeMason, 2005b; DeMason and Chetty, 2011; DeMason et al., 2013). A regional, hormonal or gene sensitivity boundary exists in the center of the leaf blade that activates downstream genes to produce either bifacial leaflets or unifacial tendrils (Lu et al., 1996; DeMason et al., 2013). Other signals provide the basis for petiole and stipule position and morphogenesis. We found that the cri allele has little effect on the number of pinna pair positions produced or on the regional boundary within the blade. These conclusions are based on the facts that numbers and types of pinnae are similar on leaves of both WT and cri. This is also true of af vs. af cri, af tl vs. af cri tl, st vs. cri st, and tl vs. cri tl. The only exception is that leaves on the cri uni-tac genotype have fewer pinna pairs than those of either uni-tac or cri, and, further, uni-tac has very low expression of CRI in its shoot tips. These observations suggest that UNI and CRI interact to regulate pinna production. Finally, leaves on the cri genotypes, except for cri uni-tac, often produce two separate pinnae (usually leaflets) at the proximal-most pinna position on leaves, suggesting a slight tendency for increased leaf complexity in the proximal region of the blade. We prefer to interpret these leaflets as doubled, often with congenital fusion, rather than as lobed as Tattersall et al. (2005) did because they were often completely separate (unfused) and on one cri st plant two different pinna types (both a compound tendril and a leaflet) were attached to the same position. The lack of pinna doubling on the cri uni-tac genotype may be the result of low UNI expression or low auxin response in the uni-tac genotype (DeMason and Chetty, 2011; DeMason et al., 2013), especially since, as we describe below, UNI expression is high in shoot tips of cri mutants. Many characteristics of the cri mutants illustrate that the CRI gene affects aspects of proximal–distal asymmetry in pea leaves and has stronger effects on the proximal region than on the distal, hinting at gradients (Table 1). This interpretation is subtly different from the one proposed by Tattersall et al. (2005) that CRI only affects the proximal region of the leaf. Focusing on the proximal leaf region, we found many anomalous features of cri mutants including, foreshortening of the petiole and proximal rachis segments, reiteration of stipule production beneath proximal pinnae and pinna doubling at the proximal-most pinna position (discussed above). Foreshortening, pinna doubling and ectopic stipules occur in the proximal regions of leaves irrespective of pinna type. These characteristics occur on af cri (which

March 2014]


421

Fig. 9. (A–F) Morphology and (G–K) anatomy of “knots” on leaflets of the cri mutant. G, H, I, J, K = morphologies that are consistent with histological sections illustrated in G–K. Insert in I is a high magnification image of vascular bundle indicated. p = palisade parenchyma, ph = phloem, x = xylem. Arrows indicate additional vascular bundles within “knots”. Scale bar in C also represents A, B, D–F; in G also represents H–K.

422


[Vol. 101

Fig. 10. GUS-staining of leaflets on DR5::GUS cri plants. mt = marginal teeth. Asterisks (*) indicate blue staining in tips of “knots.” Scale bars = 500 µm.

has branched tendrils in the proximal blade), af cri tl (which has compound pinnae in the proximal blade), cri tl, cri uni-tac, and cri (which all have leaflets in the proximal blade). Only the cri st genotype lacks distal reiteration of stipules and Tattersall et al. (2005) interpreted this to indicate that the ST gene defines the stipule in pea leaves. Our images show that the morphology of stipules on the st genotype is much more elaborate than has been reported in the past. Also, cri and st have additive effects on petiole length and stipule form, and anatomy and CRI expression is very low in st shoot tips. Therefore, we propose that CRI and ST may interact to partition the young leaf primordium into future leaf components, especially those of the leaf base (Table 1). Reiteration of stipules in the proximal region of the blade is an example of basal leaf identity displacement distally in the leaf blade. Various examples of proximal–distal displacement have been identified in ARP mutants of other species such as Antirrhinum, Arabidopsis, tomato, Nicotiana, and maize. Although cri mutants of pea and as1 mutants of Arabidopsis (Byrne et al., 2000, 2002) have shortened petioles, the petioles on ARP mutant leaves can be elongate and displace the lamina to the distal tip of the leaf. This has been demonstrated for antisense PHAN tomato plants (Kim et al., 2003; Zoulias et al., 2012). Adult leaves of antisense NSPHAN plants also have a very elongate region of unifacial morphology, and the bifacial and

lobed blade is relegated to the distal tip (McHale and Koning, 2004). Although these authors interpreted the unifacial region as being “stem-like”, based on the results in tomato, it is reasonable to interpret them as elongate, unifacial, distally extended petioles. The same interpretation could explain the adult leaf phenotype on phantastica mutants of Antirrhinum (Waites and Hudson, 1995). Various types of anomalies that result in a displacement of sheath, auricle, or ligule entities into the blade region as “knots” occur in gain-of-function KNOX maize leaves as well (Hake, 1992; Ramirez et al., 2009). We found distinct evidence that the CRI gene also affects development of the distal region of the leaf blade in pea. This is only detected in genotypes that have distal leaflets (i.e., cri unitac, cri tl, and af cri tl), where we saw that leaflets in both regions of the leaf are abnormally narrow, have abaxially recurved margins, and may also possess “knots”. Further, form abnormalities in leaflets were more common in the distal than in the proximal blade region of leaves on the cri tl and af cri tl genotypes. Some of these abnormalities resulted in unifacial or partially unifacial forms such as trumpet-shape leaflets or needle-like leaflets on the cri tl genotype (Table 1). These observations suggest that effects of cri in the distal region of genotypes with distal tendrils are not evident because tendrils are already essentially unifacial.

March 2014]


423

Fig. 11. Relative gene expression of NAM1/2, CUC3, HOP1/PsPK1, LE, UNI, TL, and PIN1 in shoot tips of WT controls (black) and the cri mutant (gray). Based on Student t tests, all expression differences were significant at 0.01 level (**) or at 0.05 level (*). ΔΔCts and results of t tests are available in online Appendix S2.

Adaxial–abaxial symmetry or bifaciality is not generally compromised in stipules or leaflets of the cri mutant despite the interpretation presented in Tattersall et al. (2005, p. 1052) that “stipules and leaflets lack adaxial identity.” Bifacial, meaning “two faces” of flattened plant organs, such as leaves, stipules, bracts, and scale leaves, is typically defined in terms of the arrangement of cell types in the ground tissue system (i.e., palisade parenchyma on the adaxial side and spongy mesophyll on the abaxial side) (Esau, 1977) and is secondarily bolstered by arrangement of cell types present in the vascular tissue system (i.e., xylem facing the adaxial side and phloem facing the abaxial side) (Cutter, 1971). Bifaciality can also be evident in the epidermis by the presence of distinctive trichomes or differences in stomatal distribution, which do not occur in pea. Stipules and leaflets on cri plants possess palisade parenchyma on the adaxial side of the mesophyll and possess vascular bundles with xylem facing the palisade parenchyma. The only flattened structure that does not possess that histological cell type arrangement in the ground tissue system is ectopic stipules. It is also somewhat indistinct in the very narrow stipules on leaves of the cri st mutant. These two structures are smaller than all other flattened leaf components. However, both do have normal bifacial vascular bundles. There is also no loss of bifaciality in “knot”-free regions of leaves on antisense NSPHAN plants of tobacco (McHale and Koning, 2004) or maize rough sheath2 mutants (Timmermans et al., 1999; Tsiantis et al., 1999). In fact, Iwasaki et al. (2013) have recently suggested that the AS1AS2 transcription complex only functions to “stabilize” adaxial– abaxial symmetry in Arabidopsis leaf development. We also investigated adaxial–abaxial symmetry of petioles. Cell shape and cell type arrangements in the petioles of cri mutants vs. WT are more divergent than stipules or leaflets. Although petioles of both genotypes are hollow, those on cri mutants are grooved on the adaxial surface and appear taller in the adaxial–abaxial direction than in width, which is not true of

Fig. 12. Relative gene expression of (A) CRI and (B) HOP1/PsPK1 in shoot tips of different pea genotypes. WT controls are in black and mutants are in gray. Significance for Student t tests between each genotype and the WT control are presented above each bar. * Significance at 0.05 level; ** significance at 0.01 level; NS indicates no significance. ΔΔCts and results of t tests are available in online Appendix S3.

WT petioles. Fusion of the stipule to the adaxial side of the petiole in cri mutants results in: (1) extreme curvature and subsequent contortions, which prevented us from taking measurements on petiole dimensions, and (2) anatomical rearrangements, which cause the petiole–stipule anatomy to resemble the anatomy typical of simple, bifacial leaves. To provide accurate information on petiole development in cri mutants, we used the cri st genotype, which lacks stipule fusion. Compared with those of WT, these petioles have significantly greater adaxial expansion during development suggesting excessive activity of an adaxial meristem. Vascular bundle anatomy is relatively circular in outline, like WT, but chlorenchyma is lacking on the adaxial side. Instead, callus-like “knots” and evidence of recent periclinal cell divisions in the cortical parenchyma are evident in most sections. These observations support the interpretation that excessive adaxial expansion occurs during petiole development in the cri mutants (Table 1). Clearly, loss of ARP gene activity has somewhat different effects on petiole development in pea than it does in tomato,

424


[Vol. 101

TABLE 1.

Role of CRI gene in pea leaf morphogenesis in three planes of symmetry when CRI gene suppresses KNOX1 and UNI genes, and mutant characters in support of these functions.

where antiSlPHAN plants have a petiole/rachis axis that is more unifacial than normal (Zoulias et al., 2012). The final plane of leaf asymmetry is mediolateral, which is most applicable to development of leaflets in the compound leaves of pea. Although Lamm (1949) observed that cri mutants possessed long leaflets and Tattersall et al. (2005) observed that they had reduced width to length ratios, we found that leaflet width, but not length, differed in the cri mutants. Consistent with reduced leaflet width, we also found that secondary vein angle is more acute, areoles are smaller, and marginal teeth are more closely spaced in cri leaflets than in WT. Together, these facts suggest that leaflet blade expansion is compromised in cri mutants (Table 1), which could be the result of (1) fewer cell divisions associated with plate meristematic activity or (2) reduced cell expansion or both. Antisense SlPHAN tomato plants, antisense NSPHAN tobacco plants, phan mutants of Antirrhinum and as1 mutants of Arabidopsis all have narrow leaves or leaflets in comparison to normal plants (Waites and Hudson, 1995; Byrne et al., 2000; McHale and Koning, 2004; Zoulias et al., 2012). Although we did not specifically measure stipule widths, they appear narrower in cri mutants as well, which is particularly evident in the cri st genotype. Cell widths as seen in sections, do not appear to differ in cri stipules or leaflets compared to WT, so we suggest that the lack of blade expansion is probably due to plate meristem cell division activity. Zoulias et al. (2012) showed that CaMV35S::SlPHAN leaflets have prolonged cell division and expansion in tomato and that the winged petiole of CaMV35S::SlPHAN tobacco plants expand more than those of WT leaves.

“Knot” morphology and anatomy—Peas, like other species in which reduced ARP gene phenotypes have been studied, possess epiphyllous “knots” on the adaxial surface of leaflets. The surface morphology of “knots” on pea leaflets varies, but “knots” are always a localized elaboration of the adaxial surface. The anatomy of pea “knots” is also variable, but it always consists of a vascularized out-pocketing of the adaxial cell surface of the leaflet with either a continuous layer of adaxial palisade parenchyma or with inverted bifacial histology. This histology is most similar to the ectopic bifacial leaf blades produced on juvenile leaves of antisense NSPHAN tobacco plants as described by McHale and Koning (2004). Although Tattersall et al. (2005) speculated that pea “knots” might resemble patches on Antirrhinum leaves where adaxial cell types are replaced by abaxial ones (Waites et al., 1998), we found no evidence for this in our sections. Instead, even inverted “knots” were always bifacial, somewhat elevated and associated with additional vascular tissue. If one includes the pea “knots” in the greater context of localized areas of excessive adaxial cell division activity, one can also consider other aspects of the cri leaf phenotype as “knot” manifestations, such as production of ectopic stipules on the adaxial surface of the rachis, callus-like growths on the adaxial surface of cri st petioles, fusion of the stipule to the adaxial side of petioles, the undulate adaxial surface of stipules/leaflets, and abaxial curling of leaflet margins (Table 1). The undulate surfaces appear to result from excessive palisade parenchyma production above vascular bundles. McHale and Koning (2004) also described abnormal proliferation of palisade parenchyma

March 2014]


over veins in leaves of antisense NSPHAN tobacco plants. Procambial strands are known to be sites of both KNOX1 and UNI gene expression. The presence of many types of ectopic adaxial expansion in pea suggests that the CRI gene functions to suppress these phenomena (Table 1). “Knots” and other forms of excessive adaxial expansion are also common in ARP mutants of other species. The magnifica form of the as1 mutant of Arabidopsis has callus-like outgrowths on leaves and ectopic shoots formed on petioles (Byrne et al., 2000). Antisense SlPHAN tomato plants produce leaves with “knots” on the adaxial surface and ectopic shoots on the midrib (Zoulias et al., 2012). “Knot” form and anatomy provides no evidence that CRI specifically defines adaxial identity in pea leaves because adaxial cell types are present in all pea “knots”. McHale and Koning (2004) also concluded this for NSPHAN and tobacco leaf development. Our interpretation is that CRI functions to suppress ectopic adaxial cell proliferation and thus possibly stabilizes adaxial–abaxial symmetry, which is primarily controlled by other genes. Gene expression and hormone involvement in CRI function in pea—In spite of the fact that UNI, instead of KNOX1 genes, primarily controls pinna pair development during morphogenesis of compound leaves in pea, there are many similarities between the cri phenotype and those of other plant specie, suggesting that some similar roles are played by the pea ARP gene. We explored gene expression differences in cri shoot tips compared with WT by investigating expression levels of a number of auxin-regulated genes (TL and PIN1) and genes known to play common roles in leaf development in species with compound leaves (NAM1/2, CUC3), LE (a gene in the GA biosynthetic pathway known to be downregulated by GA), UNI, HOP1/PsKN1, and CRI. We found no evidence that higher than normal auxin levels play a role in the morphogenesis stages of leaf development in cri since neither TL nor PIN1 have elevated expression levels and GUS expression is also not elevated in shoot tips of DR5::GUSexpressing plants. Further, we did not see evidence of GUS staining associated with “knot” initiation on leaflets. We only observed staining in “knot” tips at later stages of expansion and vascularization. If anything, auxin levels or response are probably lower in cri mutants compared with WT. It is interesting to note that ectopic stipules are present in sinuses of leaf blades in Arabidopsis as1;axr1 double mutants, suggesting that both the ARP gene, AS1, and normal auxin signaling interact to downregulate ectopic expression of the KNOX gene BP (=KNAT1), causing this parallel anomaly (Hay et al., 2006). Other auxin-regulated genes, NAM1/2 and CUC3, which have been implicated in regulation of compound leaf morphogenesis in a number of species, including pea (Blein et al., 2008) also have lower mRNA levels in cri shoot tips compared with WT. These genes are thought to define boundaries between regions of primordium initiation by local suppression of cell division activity at the margins. Blein et al., (2008) demonstrated that downregulation of these genes in pea results in fusions of the leaflets and tendrils to the rachis, and loss of distal leaflets and tendrils. However, they did not show fusions or other abnormalities in the proximal region of the leaf. We hypothesize that CRI interacts with NAM1/2 and CUC3 to suppress within leaf fusions (i.e., petiole–stipule fusions) more in the leaf base where CRI has it strongest effects. An important function of KNOX1/ARP genes is to modulate hormone levels for meristem maintenance and organ initiation, where these genes have their greatest effects. The mechanism of KNOX1 regulation of GA levels is known, and reduced GA levels occur in KNOX1 overexpressing plants (Hay et al., 2002, 2004;

425

Piazza et al., 2005; Bolduc and Hake, 2009; Bolduc et al., 2012). We looked at mRNA levels of the GA biosynthesis gene LE to assess relative GA levels in cri shoot tips because this gene is downregulated by GA. In contrast to the general pattern seen in other species, LE is reduced in cri shoot tips despite the fact that the KNOX1 gene, HOP1/PsKN1 is upregulated. Therefore, the relationship of the KNOX1/ARP module and GA levels in pea differs from that of other species since the role of GA promotes rather than inhibits leaf complexity morphogenesis, as has been established previously (DeMason and Chetty, 2011). We found that the HOP1/PsKN1 mRNA level is higher in cri shoot tips than in WT. This was an unexpected observation based on the conclusions of Tattersall et al. (2005). They localized mRNA expression in sections of WT and cri shoot tips by in situ hybridization and found that PsKN2, but not HOP1/ PsKN1 had expanded expression in the mutant leaf primordia. However, HOP1/PsKN1 expression was very low in leaf primordia and in situ hybridization is not a quantitative technique. ARP genes in other species have been demonstrated to downregulate the STM-like gene, including maize and Antirrhinum (Jackson et al., 1994; Harrison et al., 2005). In tomato, SlPHAN downregulates TKN1 (the KNAT1/BP-like gene) in both the shoot apical meristem and in leaf primordia and LeT6 (the STM-like gene) in the shoot apical meristem (Zoulias et al., 2012). We consider the regulatory relationships in pea to be somewhat equivocal. Currently, the evidence supports the hypothesis that CRI regulates both known KNOX1 genes in pea, but future quantitative and mechanistic experiments are necessary to test it. The UNI mRNA level is significantly higher in cri shoot tips than in those of WT. This observation combined with the facts that the cri uni-tac double mutant has fewer pinna pairs per leaf than uni-tac and that CRI mRNA level is very low in uni-tac shoot tips suggests an interaction between UNI and KNOX1/ ARP signaling pathways in pea. Our exploration of HOP1/ PsKN1 and CRI expression in shoot tips of the different classic leaf architecture mutants of pea does not provide any clues to the nature of this interaction because there is no correlation between differential UNI expression patterns (DeMason et al., 2013) and those of these two genes in the genotypes studied. Conclusions— As the phenotypes of ARP mutants in more plant species become known, the commonalities will provide a more accurate picture of the role that these genes play in leaf development. The many similarities we described for the cri phenotype in pea compared with those of other plant species strongly suggest that the KNOX1/ARP module plays some similar roles across plants with different leaf developmental mechanisms. We conclude that the CRISPA gene of pea suppresses KNOX1 and UNI genes and functions (1) to maintain the proximal–distal regions in their appropriate positions, thus defining the base, petiole, and blade; (2) to restrict excessive adaxial cell proliferation; and (3) to promote laminar expansion (Table 1). LITERATURE CITED BAI, F., AND D. A. DEMASON. 2006. Hormone interactions and regulation of Unifoliata, PsPK2, PsPIN1 and LE gene expression in pea (Pisum sativum) shoot tips. Plant & Cell Physiology 47: 935–948. BERDNIKOV, V. A., F. L. GOREL, V. S. BOGDANOVA, AND O. E. KOSTERIN. 2000. Interaction of a new leaf mutation ins2 with af, unitac and tlw. Pisum Genetics 32: 9–12.

426


BHARATHAN, G., T. E. GOLIBER, C. MOORE, S. KESSLER, T. PHAM, AND N. R. SINHA. 2002. Homologies in leaf form inferred from KNOX1 gene expression during development. Science 296: 1858–1860. BHARATHAN, G., AND N. R. SINHA. 2001. The regulation of compound leaf development. Plant Physiology 127: 1533–1538. BLEIN, T., A. PULIDO, A. VIALETTE-GUIRAUD, K. NIKOVICS, H. MORIN, A. HAY, I. E. JOHANSEN, ET AL. 2008. A conserved molecular framework for compound leaf development. Science 322: 1835–1839. BOLDUC, N., AND S. HAKE. 2009. The maize transcription factor KNOTTED1 directly regulates the gibberellin catabolism gene ga2ox1. Plant Cell 21: 1647–1658. BOLDUC, N., A. YILMAZ, M. K. MEJIA-GUERRA, K. MOROHASHI, D. O’CONNER, E. GROTEWOLD, AND S. HAKE. 2012. Unraveling the KNOTTED1 regulatory network in maize meristems. Genes & Development 26: 1685–1690. BYRNE, M. E., R. BARLEY, M. CURTIS, J. M. ARROYO, M. DUNHAM, A. HUDSON, AND R. A. MARTIENSSEN. 2000. Asymmetric leaves1 mediates leaf patterning and stem cell function in Arabidopsis. Nature 408: 967–971. BYRNE, M. E., J. SIMOROWSKI, AND R. A. MARTIENSSEN. 2002. ASYMMETRIC LEAVES1 reveals knox gene redundancy in Arabidopsis. Development 129: 1957–1965. CHAMPAGNE, C., AND N. SINHA. 2004. Compound leaves: Equal to the sum of their parts? Development 131: 4401–4412. CHEN, J. J., B. J. JANSSEN, A. WILLIAMS, AND N. SINHA. 1997. A gene fusion at a homeobox locus: Alterations in leaf shape and implications for morphological evolution. Plant Cell 9: 289–304. CRONK, Q. C. B. 2009. The molecular organography of plants. Oxford University Press, New York, New York, USA. CUTTER, E. G. 1971. Plant anatomy: Experimentation and interpretation, Part 2, Organs. Addison-Wesley, Reading, UK. DEMASON, D. A. 2005a. Auxin-cytokinin and auxin-gibberellin interactions during morphogenesis of the compound leaves of pea (Pisum sativum). Planta 222: 151–166. DEMASON, D. A. 2005b. Extending Marx’s isogenic lines in search of Uni function. Pisum Genetics 37: 10–15. DEMASON, D. A., AND R. CHAWLA. 2004a. Roles of auxin and Uni in leaf morphogenesis of the afila genotype of pea (Pisum sativum). International Journal of Plant Sciences 165: 707–722. DEMASON, D. A., AND R. CHAWLA. 2004b. Roles for auxin during morphogenesis of the compound leaves of pea (Pisum sativum). Planta 218: 435–448. DEMASON, D. A., AND V. CHETTY. 2011. Interactions between GA, auxin and UNI expression controlling shoot ontogeny, leaf morphogenesis and auxin response in garden pea; or how the uni-tac mutant is rescued. American Journal of Botany 98: 775–791. DEMASON, D. A., V. CHETTY, L. S. BARKAWI, X. LIU, AND J. D. COHEN. 2013. Unifoliata-Afila interactions in pea leaf morphogenesis. American Journal of Botany 100: 478–495. DEMASON, D. A., AND P. L. POLOWICK. 2009. Patterns of DR5:GUS expression in organs of pea (Pisum sativum). International Journal of Plant Sciences 170: 1–11. DEMASON, D. A., AND R. J. SCHMIDT. 2001. Roles of the Uni gene in shoot and leaf development of pea (Pisum sativum): Phenotypic characterization and leaf development in the uni and uni-tac mutants. International Journal of Plant Sciences 162: 1033–1051. DEMASON, D. A., AND P. J. VILLANI. 2001. Genetic control of leaf development in pea (Pisum sativum). International Journal of Plant Sciences 162: 493–511. DOYLE, J. 2007. Systematic value and evolution of leaf architecture across the angiosperms in light of molecular phylogenetic analyses. Cour. Forschungsinstitut Senckenberg 258: 21–37. EFRONI, I., Y. ESHED, AND E. LIFSCHITZ. 2010. Morphogenesis of simple and compound leaves: A critical review. Plant Cell 22: 1019–1032. ESAU, K. 1977. Anatomy of seed plants, 2nd ed. John Wiley, New York, New York, USA. FOSTER, A. S. 1936. Leaf differentiation in angiosperms. Botanical Review 2: 349–372. FRANCK, D. H. 1976. The morphological interpretation of epiascidiate leaves—An historical perspective. Botanical Review 42: 345–388.

[Vol. 101

FREELING, M., AND S. HAKE. 1985. Developmental genetics of mutants that specify Knotted leaves in maize. Genetics 111: 617–634. GILES, J. E., P. J. VILLANI, AND D. A. DEMASON. 1998. A class 1 knox fulllength cDNA from pea (Pisum sativum) shoot tips (accession no. AF063307). (Plant Gene Register 98-115). Plant Physiology 117: 1125. HAKE, S. 1992. Unraveling the knots in plant development. Trends in Genetics 8: 109–114. HAREVEN, D., T. GUTFINGER, A. PARNIS, Y. ESHED, AND E. LIFSCHITZ. 1996. The making of a compound leaf: Genetic manipulation of leaf architecture in tomato. Cell 84: 735–744. HARRISON, C. J., S. B. CORLEY, E. C. MOYLAN, D. L. ALEXANDER, R. W. SCOTLAND, AND J. A. LANGDALE. 2005. Independent recruitment of a conserved developmental mechanism during leaf evolution. Nature 434: 509–514. HAY, A., M. BARKOULAS, AND M. TSIANTIS. 2006. ASYMMETRIC LEAVES1 and auxin activities converge to repress BREVIPEDICELLUS expression and promote leaf development in Arabidopsis. Development 133: 3955–3961. HAY, A., J. CRAFT, AND M. TSIANTIS. 2004. Plant hormones and homeoboxes: Bridging the gap? BioEssays 26: 395–404. HAY, A., H. KAUR, A. PHILLIPS, P. HEDDEN, S. HAKE, AND M. TSIANTIS. 2002. The gibberellin pathway mediates KNOTTED1-type homeobox function in plants with different body plans. Current Biology 12: 1557–1565. HAY, A., AND M. TSIANTIS. 2009. A KNOX family TALE. Current Opinion in Plant Biology 12: 593–598. HOFER, J. M., C. GOURLAY, A. MICHAEL, AND T. H. N. ELLIS. 2001. Expression of a class 1 knotted-like homeobox gene is down-regulated in pea compound leaf primordia. Plant Molecular Biology 45: 387–398. IWASAKI, M., H. TAKAHASHI, H. IWAKAWA, A. NAKAGAWA, T. ISHIKAWA, H. TANAKA, Y. MATSUMURA, ET AL. 2013. Dual regulation of ETTIN (ARF3) gene expression by AS1-AS2, which maintains the DNA methylation level, is involved in stabilization of leaf adaxial-abaxial partitioning in Arabidopsis. Development 140: 1958–1969. JACKSON, D., B. VEIT, AND S. HAKE. 1994. Expression of maize Knotted1 related homeobox genes in the shoot apical meristem predicts patterns of morphogenesis in the vegetative shoot. Development 120: 404–413. KAPLAN, D. R. 1970a. Comparative development and morphological interpretation of ‘rachis-leaves’ in Umbelliferae. In N. K. B. Robson, D. F. Cutler, and M. Gregory [eds.], New research in plant anatomy, 101–125 [Botanical Journal of the Linnean Society 63, supplement 1]. Academic Press, London, UK. KAPLAN, D. R. 1970b. Comparative foliar histogenesis in Acorus calamus and its bearing on the phyllode theory of monocotyledonous leaves. American Journal of Botany 57: 331–361. KAPLAN, D. R. 1973a. Comparative developmental analysis of the heterblastic leaf series of axillary shoots of Acorus calamus L. (Araceae). La Cellule 69: 253–290. KAPLAN, D. R. 1973b. The monocotyledons: Their evolution and comparative biology. VII. The problem of leaf morphology and evolution in the monocotyledons. The Quarterly Review of Biology 48: 437–457. KAPLAN, D. R. 1975. Comparative developmental evaluation of the morphology of unifacial leaves in the monocotyledons. Botanishche Jahrbűcher Syst. 95: 1–105. KAPLAN, D. R. 1980. Heteroblastic leaf development in Acacia: Morphological and morphogenetic implications. La Cellule 73: 137–203. KIM, M., S. MCCORMICK, M. TIMMERMANS, AND N. SINHA. 2003. The expression domain of PHANTASTICA determines leaflet placement in compound leaves. Nature 424: 438–443. LAMM, R. 1949. Contributions to the genetics of the Gp-chromosome of Pisum. Hereditas 35: 203–214. LU, B., P. J. VILLANI, J. C. WATSON, D. A. DEMASON, AND T. J. COOKE. 1996. The control of pinna morphology in wild type and mutant leaves of the garden pea (Pisum sativum L.). International Journal of Plant Sciences 157: 659–673. MARX, G. A. 1974. Stocks available. Pisum Newsletter 6: 60. MCHALE, N. A., AND R. E. KONING. 2004. PHANTASTICA regulates development of the adaxial mesophyll in Nicotiana leaves. Plant Cell 16: 1251–1262.

March 2014]


PIAZZA, P., S. JASINSKI, AND M. TSIANTIS. 2005. Evolution of leaf developmental mechanisms. New Phytologist 167: 693–710. RAMIREZ, J., N. BOLDUC, D. LISCH, AND S. HAKE. 2009. Distal expression of knotted1 in maize leaves leads to reestablishment of proximal/distal patterning and leaf dissection. Plant Physiology 151: 1878–1888. SAINSBURY, F., A. D. TATTERSALL, M. J. AMBROSE, L. TURNER, T. H. N. ELLIS, AND J. M. I. HOFER. 2006. A crispa null mutant facilitates identification of a crispa-like pseudogene in pea. Functional Plant Biology 33: 757–763. SANDERS, H., G. W. ROTHWELL, AND S. E. WYATT. 2009. Key morphological alterations in the evolution of leaves. International Journal of Plant Sciences 170: 860–868. STEEVES, T. A., AND I. M. SUSSEX. 1989. Patterns in plant development. Cambridge University Press, Cambridge, UK. TATTERSALL, A., L. TURNER, M. R. KNOX, M. J. AMBROSE, T. H. N. ELLIS, AND J. M. I. HOFER. 2005. The mutant crispa reveals multiple roles for PHANTASTICA in pea compound leaf development. Plant Cell 17: 1046–1060. TIMMERMANS, M. C. P., A. HUDSON, P. W. BECRAFT, AND T. NELSON. 1999. ROUGH SHEATH2: A myb protein that represses knox homeobox genes in maize lateral organ primordia. Science 284: 151–153. TSIANTIS, M., R. SCHNEEBERGER, J. F. GOLZ, M. FREELING, AND J. A. LANGDALE. 1999. The maize rough sheath2 gene and leaf development programs in monocot and dicot plants. Science 284: 154–156.

427

VILLANI, P. J., AND D. A. DEMASON. 1997. Roles of the Af and Tl genes in pea leaf morphogenesis: Characterization of the double mutant (afaf tltl). American Journal of Botany 84: 1323–1336. VILLANI, P. J., AND D. A. DEMASON. 1999a. The Af gene regulates timing and direction of major development events during leaf morphogenesis in garden pea (Pisum sativum). Annals of Botany 83: 117–128. VILLANI, P. J., AND D. A. DEMASON. 1999b. Roles of the Af and Tl genes in pea leaf morphogenesis: Leaf morphology and pinna anatomy of the heterozygotes. Canadian Journal of Botany 77: 611–622. VILLANI, P. J., AND D. A. DEMASON. 2000. Roles of the Af and T1 genes in pea leaf morphogenesis: Shoot ontogeny and leaf development in the heterozygotes. Annals of Botany 85: 123–135. WAITES, R., AND A. HUDSON. 1995. Phantastica: A gene required for dorsoventrality of leaves of Antirrhinum majus. Development 121: 2143–2154. WAITES, R., H. R. N. SELVADURAI, I. R. OLIVER, AND A. HUDSON. 1998. The PHANTASTICA gene encodes a myb transcription factor involved in growth and dorsoventrality of lateral organs in Antirrhinum. Cell 93: 779–789. WEEDEN, N. F., AND D. A. DEMASON. 1999. Hop1, encoding a homeodomain-containing transcription factor, is located near Prx3 on linkage group IV. Pisum Genetics 31: 37–38. YUAN, J. S., A. REED, F. CHEN, AND C. N. STEWARD. 2006. Statistical analysis of real-time PCR data. BMC Bioinformatics 7: 85–98. ZOULIAS, N., D. KOENIG, A. HAMIDI, S. MCCORMICK, AND M. KIM. 2012. A role for PHANTASTICA in mediolateral regulation of adaxial domain development in tomato and tobacco leaves. Annals of Botany 109: 407–418.

Characterization of proanthocyanidin metabolism in pea (Pisum sativum) seeds.

Lectin-gene expression in pea (Pisum sativum L.) roots.

Isolation and characterization of novel EST-derived genic markers in Pisum sativum (Fabaceae).

Development and characterization of 37 novel EST-SSR markers in Pisum sativum (Fabaceae).

Transcription of a legumin gene from pea (Pisum sativum L.) in vitro.

Nodulation and nitrogen fixation mutants of pea, Pisum sativum.

Inheritance and mapping of isoenzymes in pea (Pisum sativum L.).

Immunolocalization of pectic polysaccharides during abscission in pea seeds (Pisum sativum L.) and in abscission less def pea mutant seeds.

Response ofNp mutant of pea (Pisum sativum L.) to pea weevil (Bruchus pisorum L.) oviposition and extracts.

Isolation and characterization of phytoferritin from pea (Pisum sativum) and Lentil (Lens esculenta).

Immunolocalization of lipoxygenase in pea (Pisum sativum L.) carpels.

Plasmatubules in transfer cells of pea (Pisum sativum L.).

Production of transgenic pea (Pisum sativum L.) plants by Agrobacterium tumefaciens - mediated gene transfer.

Isolation and characterization of NADH-glutamate synthase from pea (Pisum sativum L.).

Whole shoot mineral partitioning and accumulation in pea (Pisum sativum).

Pre-fractionation strategies to resolve pea (Pisum sativum) sub-proteomes.

Flowering time adaption in Swedish landrace pea (Pisum sativum L.).

Plant regeneration via somatic embryogenesis in pea (Pisum sativum L.).

Cloning of the gene (gdcH) encoding H-protein, a component of the glycine decarboxylase complex of pea (Pisum sativum L.).

Influence of ammonium chloride on the nitrogenase activity of nodulated pea plants (Pisum sativum).

Properties of protein-chlorophyll complexes from pea (Pisum sativum L.) leaves. The organization of chlorophyll.

The complete amino acid sequence of the alpha-subunit of pea lectin, Pisum sativum.

Isolation, characterization and partial amino acid sequence of a chloroplast-localized porphobilinogen deaminase from pea (Pisum sativum L.).

Development of the quiescent center in maturing embryonic radicles of pea (Pisum sativum L. cv. Alaska).